Food accessibility and infant birth weight in Kenya: identifying vulnerable women before and during pregnancy

Kathryn Grace, University of Utah
Molly Brown, NASA Goddard Space Flight Center

To reduce the likelihood of a woman delivering a baby with low birth weight (LBW) researchers have aimed to develop health- and education-based intervention strategies. Our aim in this study is to develop a quantitative strategy for identifying the women in Kenya - a country facing increasing food insecurity and currently coping with increasing rates of LBW - most likely to give birth to a LBW baby. In this analysis we rely on two different types of data to construct a rich data set capturing the major factors relating to LBW โ€“ the geo-referenced Kenyan Demographic and Health Survey (KDHS) data from two time periods and geo-referenced monthly food pricing data gathered by the US Agency for International Development (USAID). Using LBW as the outcome variable and a variety of variables correlated to LBW, including local pricing data โ€“ the major component of food accessibility โ€“ as independent variables, we construct classification trees to identify the subpopulations at greatest risk for giving birth to LBW babies. Through identification of these vulnerable populations in Kenya we can take an important step towards improving birth outcomes among high-risk women.

  See extended abstract

Presented in Poster Session 2